Performance Evaluation of Different Denoising Techniques for Physiological Signals

Biosignals convey information about characteristics of certain processes taking place in the living system under study. Real time biosignals are not free of noise. Noise can be defined as the unwanted modifications that a signal may suffer during capture, transmission, storage or processing. Biosign...

Full description

Saved in:
Bibliographic Details
Main Authors: Muppalla, Vineeth, Suraj, N Sri Sai Krishna, Reddy, Vyza Yashwanth Sai, Suman, D.
Format: Conference Proceeding
Language:English
Subjects:
Biosignal denoising
Discrete Wavelet Transform
Discrete wavelet transforms
Electrocardiography
Electroencephalography
IIR Digital Filter
Noise reduction
Signal to noise ratio
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Biosignals convey information about characteristics of certain processes taking place in the living system under study. Real time biosignals are not free of noise. Noise can be defined as the unwanted modifications that a signal may suffer during capture, transmission, storage or processing. Biosignals assume a great role in the diagnosis of various diseases. Corruption of biosignals with noise leads to loss of amplitude and frequency information. Eliminating the noise and recovering the original signal is a task of significance. Our work deals with the design of a filter with high performance using wavelet transformation for noise reduction based on signal to noise ratio which mitigates the baseline wandering. Two different algorithms are used for denoising the electrocardiogram (ECG) signals and electroencephalogram (EEG) signals. The former algorithm was based on the reconstruction of a noise-free signal by eliminating noise. In the latter algorithm, noise is reconstructed and subtracted from the background signal to obtain a noise free signal. Three wavelets i.e. Daubechies, Symlets and Biorthogonal mother wavelets are used for DWT. IIR digital filters, Chebyshev Type-I and Chebyshev Type-II, are also evaluated for denoising the biosignals. Denoising parameters like Signal to Noise Ratio (SNR), Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE), Percentage Root Mean Square Difference(PDR) and Root Mean Square Error (RMSE), are used to assess and compare the performances of the algorithms and the filters. Obtained results show that first algorithm performs better than the second algorithm, in terms of the denoising parameters mentioned above.
ISSN:2325-9418
DOI:10.1109/INDICON.2017.8487739